OpenAI has signed a multi-year agreement with AI chipmaker Cerebras that will provide a major boost in computing power through 2028. The companies said the deal is designed to deliver faster results for OpenAI’s customers, especially for requests that currently take longer to process. Under the agreement, Cerebras will supply 750 megawatts of compute beginning in 2026 and continuing through 2028, according to Cerebras. A person familiar with the deal told TechCrunch the contract is worth more than $10 billion, and Reuters also reported the same figure.
OpenAI said in a blog post that the new systems will speed up responses that need heavier processing. Cerebras CEO Andrew Feldman compared the shift to how broadband changed the internet, arguing that “real-time inference” will change how people experience AI by making it feel more immediate and natural. OpenAI echoed that idea, saying the added capacity will help improve speed and the quality of interactions as more people rely on AI tools in everyday tasks.
Cerebras has operated for more than a decade, but its profile rose sharply after the launch of ChatGPT in 2022 and the surge in demand for AI hardware that followed. The company builds systems around chips designed specifically for AI work and has claimed its approach can run faster than traditional GPU-based setups, including many systems built around Nvidia hardware. The agreement with OpenAI positions Cerebras as a bigger player in a market where access to compute has become one of the main limits on how quickly AI products can grow.
The deal also lands as Cerebras continues to explore a public offering. The company filed for an IPO in 2024 but has delayed it several times since then. It has kept raising money while it waits, and a report this week said Cerebras has been in talks to raise another $1 billion at a $22 billion valuation. OpenAI’s CEO Sam Altman is already an investor in Cerebras, and OpenAI previously considered buying the company.
OpenAI said the Cerebras partnership fits into a broader plan to spread its computing needs across different systems rather than relying on a single source. Sachin Katti of OpenAI said Cerebras adds a dedicated low-delay option that should lead to faster replies, smoother conversations, and a stronger base for scaling real-time AI to far more users.
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.